PARAMETERS

Bimester Season Start hour End hour Time bracket

Jan-Feb S1 0 3 T1

Mar-Apr S2 3 6 T2

May-Jun S3 6 8 T3

Jul-Aug S4 8 10 T4

Sep-Oct S5 10 12 T5

Nov-Dec S6 12 14 T6

14 16 T7

16 18 T8

18 21 T9

21 24 T10

Table 3.7: Seasons and time brackets used in the models of Aegadian islands.

Legend Year Season Time bracket

Jan-Feb Nov-Dec

Figure 3.7: Year splitting in the models of Aegadian’s islands.

method are set to 5% and to "sinking fund depreciation" respectively.

Demands

The demands have been sized according to the data collected in the study of the islands’ energy systems (see 2). These demands correspond to the commodities

"DF_RSD", "DF_NRSD", "DF_ATSE", and "DF_TRAGHETTI". For the elec-tricity demands (in MWh), the sizing has been done by taking the billed energies’

data and distributing it in the year by following the electricity production plot of the diesel power plants. For the ferries demands (in n° of routes), instead, the demand has been calculated by checking the timetables of the ferries companies and combining them with the data obtained with the methodology in the paragraph 2.3.5.

Performance

The performance parameters strongly depend on the type of the adopted technology.

Their values have been extrapolated mostly from the literature, while others have been calculated with the methodology described in the chapter 2. They are summarized in the table 3.10.

Technology Costs

In the models, the technology costs could be constant in the years, or they could vary. For most of the technologies adopted in the models, the costs are assumed to be constant, while for PV panels and wind turbines they are assumed to decrease [71]. They are summarized in the table 3.8.

Capital costs Fixed costs Variable costs

Diesel imports - - 1.77 k€/t

Diesel PP 1023.5 k€/MW 30.705 k€/MW/y 0.019 k€/MWh

PV 628₂₀₂₀ - 330₂₀₅₀ k€/MW 8.16 k€/MW/y

-Wind Turbine 13252020 - 1118₂₀₅₀ k€/MW 14.575 k€/MW/y 0.003 k€/MWh Li plant 1596 k€/MW 3.916 k€/MW/y 0.006 k€/MWh Electrolyzer 1691 k€/MW 12.91 k€/MW/y 0.0056 k€/MWh

Fuel cell 1234.4 k€/MW 11.92 k€/MW/y 0.00044 k€/MWh LAES plant 1851.2 k€/MW 46.28 k€/MW/y 0.0052 k€/MWh Diesel Ferry 5000 k€/Ferry 1472 k€/Ferry/y

-Electric Ferry 16255 k€/Ferry 1472 k€/Ferry/y

-Table 3.8: Technologies’ costs in the models of the Aegadian’s islands [72, 66, 69]

Storage

The storage systems in the models are, as already said, three: lithium battery, hydrogen and LAES. The lithium battery requires only one technology, which performs both the charge and the discharge. So when ’LI_TECH’ is in mode of operation ’1’ it charges the storage, otherwise, when it is in mode of operation ’2’ it draws energy from it and injects it in the electricity network ’ELC_SC’. The same happens with the liquid air energy storage. For the hydrogen storage a different approach was used, because of the substantial difference between the technologies related to it. It is connected to the electrolyzer ’IDR_ELET’ during the charge phase and to the fuel cell ’IDR_FC’ during the discharge phase. The storages’

state of charge is set to zero when they are installed.

Storage Lifetime [y]

LI_STO 10

IDR_STO 20

LAES_STO 30

Table 3.9: Lifetime of the storages.

Capacity constraints

The renewables power potential limits calculated in the chapter 2 need to be inserted in the model, in order to have a more realistic approach when the solver installs renewable power plants. In fact, we know for sure that the available surface for the installation of wind turbines is really limited by the actual normative, also if few laws are relaxed. For the photovoltaic power potential, instead, the assumptions were of different footprint, and the available surfaces could be larger or either smaller. Those limits are set through the ’TotalAnnualMaxCapacity’ parameter.

Because of the presence of one technology for each model of turbine, they could not be installed simultaneously, due to the presence of only one site in both the islands of Favignana and Marettimo. This means that it should be created one scenario for each model of turbine, setting the turbine’s ’TotalAnnualMaxCapacity’

to the value of the power potential measured in the chapter 2, while setting the same parameter of the other turbines to zero. Leaving the model free to invest any quantity of technology potential each year could mislead the interpretation of the results (see figure 3.8). The other technologies subject to the investment constraint are all the electricity generation technologies, with an investment limited to 2 MW/year [73] The capacities of one technology unit are set only for the wind turbines, in function of their size and for the ferries, because it is not possible to install a half ferry.

2021 2,021.5 2022 2,022.5 2023 2,023.5 2024 0

2 4 6 8 10 12

2k 4k 6k 8k 10k 12k 14k

TRAGHETTI LI_TECH IMP_DSL FV ELC_D_2_RSD ELC_D_2_NRSD ELC_D_2_ATSE DIST CEN_SEA s_LI_TECH s_IMP_DSL s_FV s_ELC_D_2_RSD s_ELC_D_2_NRSD s_ELC_D_2_ATSE s_DIST s_CEN_SEA

Comparison between Tech Capacity and Annual Activity

Year

Capacity[MW|Ferries] Activity[MWh|Routes]

Figure 3.8: Comparison between the PV capacity investment (red area) and the Diesel power plant activity (blue line) if the investment constraints are not applied, in one of the Favignana’s scenarios.

Activities constraints

The activities constraints were used in order to disable the technologies in the scenarios simulated, setting them to zero. Generally speaking, they are not built for that use, but in all the simulated scenarios it didn’t seem necessary to limit the technologies activities.

Reserve margin

The reserve margin is the excess of installed capacity in respect to the peak demand.

It has been set to 1.2 and associated to the fuel ’ELC_SC’. The technologies which contribute to the capacity increase are instead the diesel power plants.

Renewable generation target

The renewable technologies in the models are the PV panels and the wind turbines, while the tagged fuel is the generators-side electricity network "ELC_SC". The target has been set to zero.

Emissions

The renewable’s generation target, if applied, forces the system to use the tagged renewable technologies to satisfy the electricity demand, but this doesn’t reduce the emissions of the technologies disconnected from the network (vehicles, ferries, cooking in the households, etc.). In order to solve this issue, rather than a renewable generation target, it has been set a CO₂ emission limit. In fact, in this way when the system is forced to reduce the pollutants, it tries to satisfy all the demands using only CO₂ free technologies (renewable generators and electric ferries in this specific case). The emissions of CO₂ are bounded to the imports of diesel (’IMP_DSL’ and

’DSL_EXT’), in a ratio of 3.15 tons of CO₂ per ton of diesel imported [74].

Technology Input fuel Inputto activityratio Input unit

Output fuelOutputto activityratioOutput unitCapacityto activityunitUnitLifetime [y] IMP_DSL---DSL1[t]999999-- DSL_EXT---DSL_TRAG1[t]999999-- DPP1DSL[2][t]ELC_SC1[MWh]8760[(MWh/y)/MW]16 FV---ELC_SC1[MWh]8760[(MWh/y)/MW]20 EOLICO3---ELC_SC1[MWh]8760[(MWh/y)/MW]20 DISTELC_SC1[MWh]ELC_D[4][MWh]999999-- LI_TECHELC_SC1.16[MWh]ELC_SC1-8760[(MWh/y)/MW]30 IDR_ELETELC_SC2.13[MWh]---8760[(MWh/y)/MW]30 IDR_FC--ELC_SC1[MWh]8760[(MWh/y)/MW]18 LAES_TECHELC_SC1.67[MWh]ELC_SC1-8760[(MWh/y)/MW]30 ELC_D_2_RSDELC_D1[MWh]DF_RSD1[MWh]999999-- ELC_D_2_NRSDELC_D1[MWh]DF_NRSD1[MWh]999999-- ELC_D_2_ATSEELC_D1[MWh]DF_ATSE1[MWh]999999-- TRAGHETTIDSL_TRAG[5][t]DF_TRAGHETTI1[Route][5][(Routes/y)/Ferry]30 TRAGHETTI_ELELC_D[5][MWh]DF_TRAGHETTI1[Route][5][(Routes/y)/Ferry]30 Table3.10:MainparametersofthetechnologiesoftheAegadian’smodels. 1 .Thedieselpowerplantnamechangesbetweeneachmodeloftheislands:CEN_SEAforFavignana,CEN_ICELfor LevanzoandCEN_SELISforMarettimo. 2 .Theinputtoactivityratiovariesdependingontheinvolveddieselpowerplant.0.236t/MWhforSEA,0.251t/MWhfor ICELand0.2366forSELIS. 3 .Forthewindturbinesthereisnotonlyonetechnology,butoneforeachturbinemodel.Seetheparagraph3.2.2formore info. 4 .Thelossesofthedistributionnetworkvarybetweentheislands.Forexample,Favignanahasthelongernetwork,with consequenthigherlosses.Itis0.936forFavignanaand0.976forbothLevanzoandMarettimo. 5 .Theferriesparameterschangedrasticallybetweentheislands,duetothedifferenttimespentforthesingleroutes(see table3.6).